Reading 06: The Cathedral and the Bazaar

It’s pretty evident that the future of software development is moving towards the bazaar model of software development. Many companies are leveraging Agile and Scrum methodologies for software development; these methodologies emphasize iterative/frequentative development, which follows The Cathedral and the Bazaar’s Principle #7: “Release early. Release often. And listen to your customers.” Likewise, many companies are now moving towards open-source software (i.e. Microsoft), which aligns with the bazaar-style of software development where independent developers improve the functionality of the software or fix bugs, as needed.

The cathedral model of software development emphasizes that the software is built based on a central plan, and then that final, “perfect” product is shipped out to the masses. “Cathedral software” seemingly wants to get it right the first time, as ESR states that “cathedrals [are] carefully crafted by individual wizards or small bands of mages working in splendid isolation, with no beta to be released before its time.” Nowadays, “getting it right the first time” (especially with software) is quite unrealistic. With that being said, I wouldn’t say that either software development model is necessarily superior, as it depends on the situation and the type of product being built, but from my experience, I prefer the bazaar model of software development. Last spring, I was in CSE Service Projects where my team and I “built” software to help a teen patient at St. Joseph’s Hospital with a learning disability learn how to read. I put quotes around the word “built” because my team and I found an open-source Chrome extension that would read aloud the text on any website the user visited. We cloned this repository and modified the code so that the phonetic spelling of the words was read aloud to the user instead. We, then, repackaged the software and added it as a new extension in the Chrome store (we gave credit where credit was due and also checked to make sure the original developers of the software were okay with this). The point is, this would never have been possible without “bazaar software.” We knew that building our own text-to-speech software would’ve been unrealistic, as we were quite inexperienced when it came to developing text-to-speech software and the quality of the software would’ve suffered due to our inexperience; instead of having to focus on building the text-to-speech capabilities of the Chrome extension, we were able to, instead, focus our attention on building what we originally set out to build: a text-to-speech extension that read aloud (and showed) the phonetic spelling of words to the user. This aligns well with The Cathedral and the Bazaar’s Principle #2: “Good programmers know what to write. Great ones know what to rewrite (and reuse).” Reusing the text-to-speech open-source software allowed us to build quality software, as we built our text-to-phonetic-speech application on top of an existing top-notch application. This is why I believe that “bazaar software” will prove triumphant in the long-run. It’s astonishing what individuals are capable of creating when given access to the right resources.